Fitbit Overestimating Calories Burned

[suzikay74]

My BMR (using several different calculators) is about 1372.   On a day when I get about 10,000 steps through normal activity (i.e. NOT exercising, just walking around work and doing stuff), fitbit tells me I'm going about 4.5-5 miles.   That should add no more than 500 calories to my total (probably closer to 400 or 450).  Again, I haven't done any additional exercise, but fitbit somehow calculates my daily calorie burn at over 2100.    This makes the app totally unreliable for managing weight!    It's showing huge calorie deficits (and I give myself a big margin of error on calories going in) but my weight is stagnant.  

[Evelyn__020]

Yes, I've used Virtuagym and another app for a long period of time and the
results between those do differ but considerably less.
The Fitbit Charge HR gave 2500 calory burn for me yesterday which cannot be
true by any stretch of the imagination! 😉 I checked sex, age, weight in
my profile and no errors there.

[Esya]

@Evelyn__020

The point is that it does not matter what the number of calories out reported by FitBit is. It is going to make the same mistake every day (wheather that is underestimating, overestimating, a little off or a lot off). You can take the number for calories out, you have a number for calories in and your weight from the scale. With these 3 sets of data you can start adjusting where needed. That is how @Daves_Not_Here found out about the 160 calories off-set in the FitBit estimation for him.

If the weight is not going down after a few weeks of collecting data, you can either increase the calories out (despite what the starting number of FitBit was) or decrease the calories in or a mix of both. If you use the food plan you can increase the calories deficit set in that plan. So in the app it might say 1000 calories deficit, but in reality it could be 500, because the estimations are off (but they are off the same way every day). It is not about what the absolute value of a number is, but about increasing/decreasing those numbers where needed for you to succeed.

[Dominique]

---

@Evelyn__020 wrote:
are there tips how to adjust the measurements for the Charge HR so it reflects better the actual
energy use? If you do indeed have data that provides a standard, perhaps
the Fitbit can be tweaked to produce the same measurements?

---

The only way to change energy expenditure reported by Fitbit is to pretend you’re shorter/taller and/or  younger/older. It’s probably something most people wouldn’t want to do.

However, @Esya and @Daves_Not_Here still have a valid point: once you’ve determined the magnitude of the overestimation (which you can do if you log your intake and weight yourself regularly), you can still use Fitbit for your weight loss journey. Let’s say you want to go for a 750 calories daily deficit, but you know Fitbit overestimates your expenditure by 500 at your usual activity level: just shoot for a 1250 deficit, knowing that the actual deficit will "only" be 750.  

[Evelyn__020]

Good advice, thanks!

[Evelyn__020]

Yes, that sounds right. Thanks! Just wondering, could it be that the Fitbit
says 2500 and it is only 1500? Bc that is the difference I expect it is
(and that does seem quite a lot!)

[Dominique]

@Evelyn__020: my guess is 2500 may be too high, but 1500 may be too low, and your actual expenditure is somewhere in between. Difficult to say without knowing anything about you except your gender, but for total expenditure to be at 1500 for a woman, you would need to be older, petite and quite sedentary, or have a medical condition like Hashimoto (thyroid) etc.

You can use an online calculator such as this one to double check the numbers. There are five standard activity levels: if your average step count is around 5000 or less, use sedentary or light activity. If it’s around 10,000, use moderate, and if it’s 15,000 or more, use high. Let us know where you fall, but my guess is it will be closer to 2000 than to 1500.

[Evelyn__020]

Thanks! I checked and was suprised by the outcome. (I'm 49, 1.73m and 65kg)

[Dominique]

---

@Dominique wrote:

---

The only way to change energy expenditure reported by Fitbit is to pretend you’re shorter/taller and/or  younger/older. It’s probably something most people wouldn’t want to do.

---

Actually, there’s another way to mitigate the impact of a wrist-based Fitbit that overestimates calories burned, if you’re wearing it on your dominant arm (which is probably not the case, but who knows): put it on your non-dominant arm.

For the past three years, I’ve been wearing 2-3 Fitbits simultaneously (each connected to a separate account) and I’ve noticed the ones on my dominant hand consistently give me more steps and more calories burned, even though there’s a setting in your profile that is supposed to take this (dominant vs. non-dominant) into consideration.

I previously had a Surge on my left arm (non-dominant) and an Alta HR on my right arm. I always hit 10,000 steps on my Alta HR well ahead of my Surge. Now that I recently got an Ionic, I promoted it to main tracker (left arm) and switched Surge to right arm: lo and behold, they now vibrate for 10k almost at the same time.

[Dominique]

OK, so your calculated BMR alone is 1325, and even if you were very sedentary, you would probably expend more than 175 on top of that. Btw, the calculator uses the same equation (Mifflin St Jeor) as Fitbit. How many steps have you been getting in average since you got your Charge 2?

[Daves_Not_Here]

@Evelyn__020 - If you want to calibrate Fitbit to remove the systemic error, you could experiment with creating and logging a daily custom "food" or "activity" named "Bias" that would adjust the caloric deficit appropriately.

The challenge is the error is not constant over time.  (For example, if my BMR starts to slow due to a "starvation response", the predicted caloric deficit would become too large and I'd need to adjust the calories associated with Bias.  One of the reasons why I like to re-estimate this error is to understand if my metabolism is slowing (it's not, yet)).

Here is the error for me over time  -- Fitbit underestimates my calorie burn from 130 to 185 calories per day.

[Evelyn__020]

I checked; it says I should have an intake of about 2000 to maintain my weight. I do find that hard to believe, to be honest? (woman, 49, 1.74m, 65kg) The Fitbit today gave me over 2800 burned!

The advise to add mock calory intake I'm now following. Pity, I love to have the option to manipulate the (speed at which the) calories burn instead.

Maths are not my best quality 😉 but could I change the stride from 71cm to, for instance 100cm? 

[Daves_Not_Here]

@Evelyn__020 -- regarding tweaking the numbers, here are my suggestions from the perspective of an amateur math hobbyist:

1. Carefully capture at least 30 days of daily "raw" data before making any adjustments (in stride length, phantom food additions, etc.).  This means daily logging of your weight and ALL the food you eat into the Fitbit app, and wearing your Fitbit device as much as possible to capture all your exercise and activity.  (Very important to capture ALL food and activity - if you forget a snack and remember it later, then enter it.  If you can't figure out the quantity of some food, then enter your best guess - entering something is better than a zero)
2. Link Fitbit to TrendWeight you can see your actual weight trend with the water weight fluctuations removed.
3. Once you have 30 days of data, download it into a spreadsheet (you can do this in the Settings section of the Dashboard).  Determine your total calorie deficit, divide it by 3,500 to determine how much weight you should have lost, compare it to your actual loss in Trendweight, and determine the daily calorie deficit error.  (NOTE:  at some point, I will post a detailed description of how to perform this step)

Once you do this, you will have a pretty good idea as to how accurately Fitbit is measuring your calorie deficit, and will be in a better position to make adjustments.  I predict that you will discover that Fitbit is closer to reality than you might have guessed.

[Dominique]

@Daves_Not_Here: I think your rational mind would appreciate the following podcast with Dr. Kevin Hall (if you want to listen to something on the subject of nutrition while commuting to or from work):

The Physiology of Fat Loss, Weight Regain & Carb or Fat Restriction

It’s kind of related to what you wrote in your post. I listened to it during my daily walk and found it very interesting.

[Daves_Not_Here]

@Dominique - thanks for the kind words and link suggestion -- I'll listen to it today.

[Daves_Not_Here]

@Dominique - thanks again for the link, which I was able to listen to.  At this point, I only have a half-baked reaction to it.  I'll admit it runs contrary to my confirmation bias - that bias having nothing to do with low-carb versus low-fat advocacy, but rather how established orthodoxies deal with heretics.

Hall strikes me as a defender of the orthodoxy and he and his interviewer seem to fairly snicker as they present results that contradict one of Taubes' hypotheses.  And without acknowledging that Taubes through NuSI has provided a platform to facilitate a true clash of competing ideas, let the chips fall where they may, a favor the orthodoxy would never return.  Of course, you can't really blame them for being sore as a boil given that Taubes' position is "not only does the emperor have no clothes, he is wholly illegitimate and needs to be assassinated."

I say my response is half-baked because I'm peripherally aware of leadership turnover at NuSI, but don't really know enough about it to revise my assumptions or draw any conclusions.  As I learn more, I'll post to this thread.

[keengkong]

I adjust my height such that it shows me as being 5 inches shorter than I am in real life.  In that way, if I wear a Fitbit Flex or a Fitbit Blaze with the heart rate feature turned off, I get the same results as I would if I wore the Fitbit One using my correct height.  I prefer not to wear the One because then I have to worry about it falling off and being uncomfortable at night.  I tested these comparisons by using a second phone and a second Fitbit account.  I believe the Fitbit One is more accurate, however, it would take extremely careful tracking for a few months to verify a device is accurate.